I don't think the world is helped by having two different imperfect heuristics, and I do think it will cause bug reports for us.
I also think if anyone wants sensible behavior (or more control) they should probably use pgrep/pkill instead of killall... (Where the multi argument semantics are already consistent across implementations.) (Though that reminds me that our pgrep/pkill -a behaves differently to the others, and -a doesn't show up in the help. But that hasn't broken anyone's scripting that I'm aware of yet, so that can wait for another day...) On Sat, Dec 16, 2017, 17:40 Rob Landley <[email protected]> wrote: > On 12/16/2017 06:59 PM, enh wrote: > > That's my point: there's no way to implement this correctly, so the > > least worst choice is to just implement it the same as everyone else. > > Whatever we do will be surprising in some instances, but "the same > > surprises as the other systems" is at least manageable. I don't see how > > anyone wins from toybox having a different but also broken heuristic. > > Grumble grumble wanna decent spec. > > Sigh. I'm annoyed by having to read two different files to do one thing > but if you feel strongly about it... > > I still need to fix the basename() stuff and add lots of tests. I'll try > to get something checked in tomorrow. > > Rob >
_______________________________________________ Toybox mailing list [email protected] http://lists.landley.net/listinfo.cgi/toybox-landley.net
